AI is now deeply embedded in business operations, driving automation, insight, and decision-making across workflows and systems. As part of their evolving partnership, SAP and Google Cloud are enabling the next wave of enterprise AI by contributing to the Agent2Agent (A2A) interoperability protocol. This initiative lays the groundwork for AI agents to securely collaborate and interact across platforms.
In addition to this milestone, SAP is also expanding access to Google’s Gemini models via its generative AI hub on SAP Business Technology Platform (SAP BTP), and integrating Google’s video and speech intelligence tools to support multimodal retrieval-augmented generation (RAG) for enterprise learning and knowledge discovery.
Together, these efforts underscore a joint vision for enterprise-ready AI that is open, flexible, and deeply grounded in business processes.
Enabling Cross-Platform Agent Collaboration
The future of work is increasingly agentic, with businesses deploying AI agents to perform real tasks—resolving customer issues, handling approvals, and collaborating across departments. SAP’s Joule architecture supports these agentic workflows across the SAP Business Suite.
However, true value lies in enabling these agents to collaborate beyond the confines of a single vendor’s ecosystem. They must securely exchange information and coordinate across different systems. The A2A protocol was created to meet this need, moving beyond basic API integrations by establishing an open standard for seamless agent interaction.
SAP has joined Google Cloud and other leaders as a founding contributor to the A2A protocol. This standard allows agents from different vendors to share context and work together, enabling automation across disconnected platforms.
Take, for example, a billing inquiry received via Gmail. Rather than switching between tools, a representative can invoke Joule directly from the email. Joule, acting as the orchestrator, initiates a dispute resolution process and engages a Google agent that pulls data from BigQuery. Together, the agents analyze the issue and provide a resolution—streamlining processes and improving efficiency.
This is the type of collaboration A2A is designed to support: AI agents working in concert to reduce friction, accelerate outcomes, and free people to focus on strategic tasks. It reflects SAP’s broader vision for Joule as an agent orchestrator—interoperable, proactive, and deeply connected to business context.
Expanding Access to Google Gemini Models
SAP continues to prioritize openness and customer choice by expanding support for Google’s models within its generative AI hub, a core part of the AI Foundation on SAP BTP.
Customers now have enterprise-grade access to Google Gemini 2.0 Flash and Flash-lite models, in addition to the existing Gemini 1.5 options. These models are optimized for high performance and low latency, enabling enterprise-grade generative AI capabilities within a secure SAP environment.
This integration allows customers to create and scale AI-driven solutions using leading models while benefiting from SAP’s contextual understanding of enterprise processes. The result is AI that’s not only powerful, but also practical, secure, and aligned with how businesses operate.
Advancing Multimodal Intelligence with Video and Speech AI
SAP is also progressing multimodal RAG capabilities, a top request from customers seeking better learning and support experiences through video.
Multimodal RAG integrates data types like text, images, audio, and video to improve retrieval and generation of insights. SAP uses Google Video Intelligence for detecting on-screen text, and Google’s Speech-to-Text API to transcribe spoken audio. The outputs are timestamped and indexed, making it possible to retrieve specific video segments quickly and accurately.
This approach allows users to search and access the exact part of a video that answers their question—transforming training materials and support resources into intuitive, highly accessible tools.
Miku Jha, Director of AI/ML and Generative AI at Google Cloud, commented, “As agentic AI evolves, seamless handling of multi-modal data—text, voice, enterprise videos, and images—becomes paramount. This introduces significant challenges for agent interoperability. An open protocol like A2A is therefore indispensable, providing the necessary framework and flexibility for agents to effectively communicate and collaborate across these diverse modalities.”
This highlights how SAP and Google Cloud are working together to make unstructured content more valuable and enhance knowledge delivery across the enterprise.
A Shared Vision for Enterprise AI
These developments underscore the strategic alignment between SAP and Google Cloud. From shaping emerging standards for agent collaboration to offering model choice and unlocking multimodal insights, both companies are committed to advancing enterprise AI that is open, composable, and grounded in business context.
